> Summarizing all the changes between P4 and P5 is not so easy,
> especially when P5 is still a moving target! Here's a short list of
> some key differences: we'll try to produce a more detailed list later
> when things have settled down a bit. There is some discussion of the
> rationale and overall design goals for P5 on the website
> (http://www.tei-c.org/P5/) which also needs bringing up to date... and
> you can always find out more by coming along to the TEI Members Meeting
> or any of the various other presentations we've been making on the
> subject over the last couple of years (for a recent example, see
> http://www.tei-c.org/Talks/2005/AWCH/index27.html et seq)
>
> 1. P5 is "natively" XML: it uses a number of new XML-related features
> which were not available or well understood at the time Ps 1 to 4 were
> created. Specifically,
> -- it uses namespaces. A p5 document must specify the namespace it
> belongs to
> -- it uses Unicode. The P4 WSD has been completely rethought, and a
> new
> element is available to represent characters not available in Unicode,
> or specific glyph variants
> -- it uses attributes from the XML namespace to represent concepts
> previously represented using the global attributes lang and id
>
> So, a P4 document which looked like this
>
> <TEI.2>
> <teiHeader>....</teiHeader>
> <text id="T1">
> <p lang="fr">Une esquisse françl;aise</p>
> </text>
> </TEI.2>
>
> should at P5 look like this
>
> <TEI xmlns="http://www.tei-c.org/ns/1.0">
> <teiHeader>....</teiHeader>
> <text xml:id="T1" >
> <p xml:lang="fr">Une esquisse française</p>
> </text>
> </TEI.2>
>
> 2. At P5 we have completely overhauled the linking facilities. The old
> distinction between internal and external links (<ptr> and <xptr>) has
> been removed. Link locations are expressed using Xlink rather than the
> TEI's extended pointer syntax. (this is less traumatic than it sounds,
> since the former is largely derived from the latter).
>
> So, where at P4 you might have written
>
> <div id="sec42"><head>Concerning identifiers</head>
> <p>The section about identifiers (<ref target="sec42">this one</ref>)
> cites
> <xref doc="someUrl" from="id (sec42)">something on the web</xref>....
>
> (having remembered to declare
> <!ENTITY someUrl SYSTEM "http://www.somewhere.net">
> previously)
>
> At P5, you would instead write
>
> <div xml:id="sec42"><head>Concerning identifiers</head>
> <p>The section about identifiers (<ref target="#sec42">this one</ref>)
> cites
> <ref target="http://www.somewhere.net#sec42">something on the
> web</xref>....
>
> 3. The way in which a TEI schema is constructed has been fundamentally
> changed. Instead of combining and modifying pre-declared sets of SGML
> declarations, using SGML syntax, you now write an XML document (an
> ODD) which references the set of TEI modules you want and documents any
> changes you want to make to them. This set of XML declarations is then
> processed to produce a schema in RelaxNG or W3C Schema or a good ole
> fashioned DTD (if you insist) . The ODD language is itself part of the
> Guidelines, and we use it to produce them. We're working on a user
> guide to the powerful new facilities this offers: see a draft at
> http://www.tei-c.org/Drafts/edw88.xml
>
> If you have already put some effort into the customization of TEI P4
> (using extension files etc.) then you may find you have quite a lot of
> work to do to get an equivalent P5 schema. On the other hand, you will
> find the new customization tools *much* easier and more powerful; and
> you may well find that things you had to do at P4 are no longer
> necessary.
>
> 4. There is a whole new module in P5 for the detailed description and
> cataloguing of manuscript sources; there are also a few new elements in
> the core module, notably <choice> for representing situations where the
> encoder wishes to present two (or more) equivalent ways of encoding a
> phenomenon such as correction. So for example, where in P4 you might
> have written
>
> ... and <sic corr="he talked" resp="Theobald">a table</sic> of green
> <reg orig="feelds">fields</reg>...
>
> at P5 you will have to write
>
> ... and <choice><sic>a table</sic><corr resp="Theobald">he
> talked</corr></choice>
> of green
> <choice><orig>feelds</orig><reg>fields</reg></choice>...
>
> A number of other minor changes are likely along similar lines. These
> form part of a general trend towards making the usage of attributes
> across the TEI scheme more internally self consistent, according to a
> set of principles discussed at the TEI Council meeting just concluded
>
> 5. The chapter on feature structure representation has been revised and
> refocussed to better serve its chief user community; it is currently
> going through the process of ISO standardization.
>
> But the best way to find out the differences is... try it! As posted
> earlier, the basic building blocks of P5 are now all in place and
> available for testing. We want to hear from you!
>
> Lou
>
>
>
> On 29 Apr 2005, at 17:32, Young, John T wrote:
>
>> This is a multi-part message in MIME format.
>>
>> ------_=_NextPart_001_01C54CD9.18E52148
>> Content-Type: text/plain;
>> charset="iso-8859-1"
>> Content-Transfer-Encoding: quoted-printable
>>
>> Could I just add my vote to Adriaan's?
>> =20
>> John Young
>> The Newton Project
>> Imperial College London
>>
>> ________________________________
>>
>> From: TEI (Text Encoding Initiative) public discussion list on behalf
>> of =
>> Adriaan van der Weel
>> Sent: Fri 29/04/2005 10:56
>> To: [log in to unmask]
>> Subject: Fwd: Advice on P4 -> P5 move
>>
>>
>>
>> I, for one, would be very grateful indeed for such a summary of
>> differences.
>>
>> Adriaan van der Weel
>> University of Leiden
>>
>>
>>
>>
>> -
>
>
>
>
From the Macmini at Burnard Towers
|
